Struct ReloadOutcome 

pub struct ReloadOutcome {
    pub summary: String,
    pub had_warnings: bool,
    pub details: Vec<String>,
    pub had_errors: bool,
    pub project_trusted: Option<bool>,
}

The outcome of a reload: a human-readable status line for the transcript (counts of what reloaded), and whether any load diagnostics appeared.

Fields

summary: String

One-line summary for the transcript note (e.g. “Reloaded 2 plugin(s), 5 skill(s), 1 prompt(s).”).

had_warnings: bool

True iff at least one extension load warning fired (ABI mismatch / skip).

details: Vec<String>

Formatted diagnostic lines (package / skill / prompt-template warnings) that the caller should surface in its UI. Kept out of eprintln! because the TUI owns the terminal: direct stderr writes during a run would corrupt the alternate-screen input row.

had_errors: bool

True iff the reload itself failed hard (e.g. the extension build or settings reload errored) and the previous resources were kept. Soft diagnostics (skill shadowing, name validation) do not set this.

project_trusted: Option<bool>

The project-trust policy this reload actually resolved and applied. None when the reload failed hard and kept the previous resources, so consumers (e.g. theme discovery) should stay on the snapshot they already match instead of re-reading a stale startup bool.